The effect of triazolam and flunitrazepam--two benzodiazepines with different half-lives--on breathing during sleep.

We performed a double-blind single-dose placebo/hypnotics crossover study randomized within groups to test the potential problems that a group of normal subjects, including subjects who snore, may face using hypnotic medications. Two benzodiazepine hypnotics--triazolam, 0.25 mg, and flunitrazepam, 2 mg tablets--were considered. Subjects were monitored with nocturnal polysomnography, including esophageal pressure (Pes) monitoring as a measure of respiratory efforts, and were given daytime performance tests. Results were analyzed for the total nocturnal sleep period and also by thirds of the night in consideration of the different half-lives of the studied drugs. Three specific respiratory variables were evaluated: mean breathing frequency for selected unit of time, "Delta Pes" (esophageal pressure at peak end-expiration minus Pes at peak end-inspiration) expressed in cm H2O, and the ratio of Delta Pes/Delta TI (inspiratory time), taken as an index of respiratory drive calculated for each respiratory cycle. There was no significant increase in either the respiratory disturbance index or the oxygen desaturation index (number of drops in arterial oxygen saturation of 4% or more per hour of sleep, as measured by pulse oximetry). There was a significant increase in mean breathing frequency with flunitrazepam compared with placebo, as well as a significantly larger percentage of time during sleep with Delta Pes above 10 cm H2O (taken as a cutoff point for normal respiratory effort) with both triazolam and flunitrazepam compared with placebo. These respiratory changes, even if significant, were minor but may become a liability in association with specific abnormalities.